US6021334A  Method for transmission by a base station equipped with a multielement antenna to a mobile  Google Patents
Method for transmission by a base station equipped with a multielement antenna to a mobile Download PDFInfo
 Publication number
 US6021334A US6021334A US08965501 US96550197A US6021334A US 6021334 A US6021334 A US 6021334A US 08965501 US08965501 US 08965501 US 96550197 A US96550197 A US 96550197A US 6021334 A US6021334 A US 6021334A
 Authority
 US
 Grant status
 Grant
 Patent type
 Prior art keywords
 matrix
 frequency
 vv
 antenna
 vector
 Prior art date
 Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
 Expired  Lifetime
Links
Images
Classifications

 H—ELECTRICITY
 H04—ELECTRIC COMMUNICATION TECHNIQUE
 H04B—TRANSMISSION
 H04B7/00—Radio transmission systems, i.e. using radiation field
 H04B7/02—Diversity systems; Multiantenna system, i.e. transmission or reception using multiple antennas
 H04B7/04—Diversity systems; Multiantenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
 H04B7/06—Diversity systems; Multiantenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station
 H04B7/0613—Diversity systems; Multiantenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station using simultaneous transmission
 H04B7/0615—Diversity systems; Multiantenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station using simultaneous transmission of weighted versions of same signal
Abstract
Description
The present invention relates to a method for transmission by a base station equipped with an antenna having a plurality of elements, or "multielement" antenna, to a mobile.
It applies, in particular, to the field of mobile radio communications, and to a method for sending a digital signal between a base station and a specific mobile in the presence of interference sources and background noise.
The invention relates more particularly to sending over the socalled down link, that is to say the transmission of a digital signal from the base station to the mobile. For this purpose, the invention uses data obtained from sending over the socalled up link relating to the same base station, that is to say from signals which are received by the multielement antenna of this station and originate from the mobile and the interference sources.
Throughout the rest of the text, "frequency" indicates "carrier frequency", "antenna" indicates "multielement antenna of the base station", "transmission" and "reception" respectively indicate "transmission by the antenna" and "reception by the antenna" and "sending" indicates "transmission and/or reception".
"Useful mobile" indicates "mobile to which the method according to the present invention is applied" and "interference source" indicates "any factor which makes a contribution to the component representing the noise in the digital signal which is sent, with the exclusion of the background noise". For example, a mobile other than the useful mobile may constitute an interference source for the useful mobile.
"Frame" indicates "sequence of consecutive samples of a signal, sufficient in number to allow calculation of the required statistical data" (these statistical data will be explained below).
Mobile radio communications is currently growing at such a rate that there is a desire to increase the number of users served simultaneously in a given radio communications network. The approach generally adopted for this is to optimize the use of the spectrum of the available transmission and reception frequencies.
In a radio communications network of the cellular type, it is, in particular, possible to allow a plurality of mobiles to communicate simultaneously by allocating them the same frequency in the same cell of the network: this is the object of the technique referred to as SDMA (Space Division Multiple Access).
In this case, use is generally made of an antenna whose radiation diagram has at least one lobe. The antenna creates energy minima in reception and transmission to the mobiles, other than the useful mobile, which share the same frequency with it and constitute interference sources for this useful mobile.
Signal processing methods, applied to multielement antennas, are known which make it possible to improve the reception by acting on the up link.
However, the steps and the parameters involved in the known methods generally depend on the nature of the propagation channels observed on the various elements of the antenna. However, these channels are themselves, in particular, dependent on the carrier frequency. When the down link uses a frequency different from that of the up link, the weightings calculated and applied to the signals received by the various elements for the up link cannot generally be reused for the down link.
In an article entitled "Adaptive transmitting antenna methods for multipath environments", Globecom'94, pages 425429, D. GERLACH and A. PAULRAJ describe a method for spatial filtering in transmission, applied to a multielement antenna. This method has a number of limitations and drawbacks.
Firstly, it assumes that there is no intersymbol interference, which is not necessarily the case in practice.
Furthermore, in order to have information on the nature of the propagation channels of the down link, this prior method requires feedback from the mobile in question, that is to say the base station periodically sends test signals to the mobile, which measures them and sends the result of the measurement back to the base station. The presence of a delay between the feedback and the retransmission by the base station imposes some degree of time stability on the quantities which are measured. The number of measurements to be sent as feedback increases if the nature of the propagation channels changes rapidly. Thus, the amount of feedback which is needed may be extremely high. Even if attempts are made to reduce it, this amount of feedback necessarily limits the rate of useful information sent.
The object of the present invention is to overcome the aforementioned drawbacks.
More particularly, one object of the present invention is to improve the transmission by suppressing the general interference level during sending, by reinforcing the relative influence of the energy transmitted by the antenna to the useful mobile, and by limiting the relative influence of the energy transmitted by the antenna to the interference sources.
In particular, the present invention has two applications to a cellular mobile radio communications network. On the one hand, in an urban environment, the invention makes it possible to increase the rate of reuse of the frequencies over the cells as a whole, which makes it possible to increase the number of network users served simultaneously, by virtue of the reduction in the general level of interference. On the other hand, the invention makes it possible to increase the range of the antenna. As a consequence, in a rural environment, the invention makes it possible to limit the number of base stations needed to cover a given region.
In order to achieve the object mentioned above, the present invention proposes a method for transmitting a digital signal composed of successive frames of samples, by a base station equipped with a multielement antenna to a specific mobile, in the presence of interference sources and background noise, with the aid of at least one reception carrier frequency and at least one transmission carrier frequency, according to which:
prior to sending:
(a) for each reception carrier frequency, a reception calibration table is developed, representing the variation in contribution, as a function of the reception direction, of the various reception elements at the said reception carrier frequency;
(b) for each transmission carrier frequency, a transmission calibration table is developed, representing the variation in contribution, as a function of the transmission direction, of the various transmission elements at the said transmission carrier frequency;
(c) at least one frequency transposition operator is calculated which approximately transforms one said reception calibration table into one said transmission or reception calibration table;
then, during sending:
(d) statistical data are calculated on the basis of a plurality of samples of a plurality of frames of the signals received by the various elements, originating from the mobile and the interference sources;
(e) for the said mobile, an optimum set of spatial weightings is calculated on the basis of the said statistical data, the said frequency transposition operator or operators, and a criterion for reinforcing the useful signal and reducing the interference sources;
(f) the contributions to the said digital signal to be transmitted by each element are respectively weighted by the weightings obtained on the basis of the spatial weightings of the said optimum set; and
(g) the said digital signal thus weighted is transmitted.
Other characteristics and advantages of the present invention will emerge from reading the following description of particular embodiments which are given by way of nonlimiting examples.
The description refers to the single FIGURE which accompanies it and which constitutes a flow chart summarizing the successive steps of the method according to the present invention, in one particular embodiment.
The case of a base station equipped with an antenna having N elements will be considered below.
The term "reception directional vector" (or "transmission directional vector") is used to denote a column vector having N components in which the m^{th} component represents the signal which would be received (or transmitted) by the m^{th} element, m varying from 1 to N, in the case of receiving (or transmitting) a plane wave of given frequency coming from (or transmitted in) a direction defined by a given angle.
As mentioned in the introduction, the method according to the present invention uses data obtained from sending over the up link, that is to say in reception. It is the up link which will be referred to below.
S.sub.λr (α) denotes the reception directional vector relating
to the direction defined by the angle α, and
to the carrier wavelength λ=c/f, in which c denotes the velocity of light and f denotes the carrier frequency.
For example, in the case of a linear antenna having omnidirectional elements of unit gain distributed uniformly with a separation d, ##EQU1## in which exp denotes the exponential function and j denotes the complex number satisfying j^{2} =1.
Let X(t) be a column vector having N components in which the m^{th} component represents the signal received at time t from the mobiles by the m^{th} element of the antenna, for m varying from 1 to N, this signal being transferred to the baseband.
It is assumed that P mobiles are each transmitting a message s_{k} (t) which reaches the antenna via multiple paths of directions α_{k},i, propagation delays τ_{k},i and complex amplitudes amp_{k},i at a frequency f=c/λ.
Then ##EQU2## in which i denotes the index of the paths assigned to the mobiles, in which amp_{k},i =AMP_{k},i.exp(2πjft_{k},i),
in which AMP_{k},i denotes the modulus of the complex amplitude amp_{k},i, and
in which the term exp(2πjft_{k},i) originates from the fact that the signal X(t) is in the baseband.
It can be seen that X(t) depends on frequency through the directional vectors S.sub.λ (α) and through the phases of the complex amplitudes amp_{k},i exclusively.
For linear modulations, each message s_{k} (t) is of the form ##EQU3## in which the coefficients a_{k},n represent the transmitted symbols, the h_{k} represent the impulse responses of the transmission/reception equipment filters and T denotes the duration of a symbol.
Let X(t)=[x_{1} (t), . . . ,x_{N} (t)]^{t} in which (.)^{T} denotes the transpose matrix and in which x_{m} (t), for m varying from 1 to N, denotes the m^{th} component of X(t).
For m varying from 1 to N, then ##EQU4## in which g_{k},m denotes the impulse response of the combination of the equipment filters and the multipath propagation channel between the k^{th} mobile and the m^{th} element of the antenna.
When the data are sampled, for example at a symbol rate of 1/T in a particular embodiment, then this gives, for each sampling time lT, in which l is an integer, ##EQU5##
It can be seen that the received data sampled on an element are the sum of the contributions of the various mobiles, each contribution being a version, filtered by a digital channel, of the symbols transmitted by the various mobiles. The number of terms in a summation over n, reduced by 1, represents the length of the intersymbol interference relating to the k^{th} mobile.
Let R_{xx} =X(t).X^{+} (t), in which (.)^{+} denotes the conjugate transpose matrix and in which t describes a set of samples of a frame of the digital signal.
Let R_{xx} be the mean of the matrices R_{xx} over a set of M, not necessarily consecutive, frames received by the antenna, M being small enough for the angles of arrival, on the various elements of the antenna, of the multiple paths originating from the mobile to be stable.
It is easy to show, assuming that, during the period in which the mean R_{xx} of the matrices R_{xx} is taken,
(i) the angular characteristics of the propagation do not vary;
(ii) the arguments of the complex amplitudes amp_{k},i vary randomly in the interval [0,2π];
(iii) the moduli of the complex amplitudes amp_{k},i, that is to say the energy of the multiple paths, do not vary;
(iv) the variations in the delays of the multiple paths are negligible compared with the duration T of a symbol;
that the mean matrix R_{xx} then converges to a matrix ##EQU6## in which A=amp_{k},i ^{2}.s_{k} (lT.sub.τk,i)^{2},
in which .^{2} denotes the square of the modulus of a complex number,
and in which E denotes the mathematical expectation.
The matrix E(X(lT).X^{+} (lT)) is independent of the frames in question. The matrix R_{xx} is assumed to be an estimate thereof.
It can be seen that taking the mean of R_{xx} over a suitable number of frames has the effect of making the phase terms of the complex amplitudes amp_{k},i of the multiple paths vanish. Assuming, in addition, that the powers amp_{k},i ^{2} of the multiple paths are independent of frequency, the matrix R_{xx} then depends on frequency through the directional vectors S.sub.λ (α_{k},i) only. This property of the matrix R_{xx}, calculated on the basis of a plurality of samples of a plurality of frames, makes it possible to employ the frequency transposition operators which were mentioned above and are described in more detail below.
The term "calibration table" is used to describe the set of directional vectors relating to a given antenna geometry.
In the rest of the text, a digital signal composed of successive frames of samples will be considered. The method according to the present invention consists in transmitting this signal by a base station equipped with a multielement antenna to a useful mobile, in the presence of interference sources and background noise. It is assumed that the up link uses at least one frequency, referred to as the reception carrier frequency, and that the down link uses at least one frequency, referred to as the transmission carrier frequency.
As shown by the single FIGURE, calibration tables are developed before sending.
For each reception carrier frequency, a reception calibration table is developed, representing the variation in contribution, as a function of the reception direction, of the various reception elements at the said reception carrier frequency.
In one particular embodiment, in order to develop each reception calibration table, a matrix is formed in which each column is a directional vector in which the m^{th} component represents the signal which would be received by the m^{th} element, m varying from 1 to N in which N is the number of elements, in the case of receiving a plane wave of frequency equal to the said carrier frequency and originating from a direction defined by a predetermined angle intrinsic to the said directional vector.
For each transmission carrier frequency, a transmission calibration table is developed, representing the variation in contribution, as a function of the transmission direction, of the various transmission elements to the said transmission carrier frequency.
In one particular embodiment, in order to develop each transmission calibration table, a matrix is formed in which each column is a directional vector in which the m^{th} component represents the signal which would be transmitted by the m^{th} element, m varying from 1 to N in which N is the number of elements, in the case of transmitting a plane wave of frequency equal to the said carrier frequency in a direction defined by a predetermined angle intrinsic to the said directional vector.
Next, a respective correction factor is applied, if appropriate, to each element of the calibration tables which are obtained, in order to take account of various characteristics of instruments contained in the transmission chain and in the reception chain. This set of correction factors may thereafter be updated periodically on the basis of the change in a plurality of measured physical parameters.
As indicated by the FIGURE, the next step in the method according to the invention consists in calculating one or more linear matrix operators, referred to as frequency transposition operators. It can in fact be shown that there is a linear operator which approximately transforms a reception calibration table into a transmission or reception calibration table.
The approximation which is used may be the least squares approximation or any other suitable approximation.
Then, during sending, as shown by the FIGURE, statistical data, on the up link, are calculated on the basis of a plurality of samples of a plurality of frames of the signals received by the various elements, originating from the mobile and the interference sources.
These statistical data are advantageously of order 2. However, they may be of higher order.
In a first particular embodiment, employing a single reception carrier frequency f_{1} and a single transmission carrier frequency f_{2}, the step of calculating the frequency transposition operators consists in calculating a single frequency transposition matrix operator T_{f1},f2 which transforms the reception calibration table C_{1} associated with the frequency f_{1} into the transmission calibration table C_{2} associated with the frequency f_{2}.
In the same particular embodiment, the step of calculating the statistical data consists for each frame of a set of M, not necessarily consecutive, frames received by the antenna, M being small enough for the angles of arrival, on the various elements of the antenna, of the multiple paths originating from the mobiles to be stable:
in calculating a matrix R_{xx} ^{f1} =X_{f1} (t).X_{f1} ^{+} (t),
in which X_{f1} (t) is a vector having N components in which the m^{th} component represents, at the carrier frequency f_{1}, the signal received at time t from the mobiles by the m^{th} element of the antenna, for m varying from 1 to N,
in which (.)^{+} denotes the conjugate transpose matrix, and
in which t describes a set of samples of the said frame, and
in estimating a matrix R_{vv} ^{f1}, either on the basis of the contributions of the interference sources and the background noise on each of the N elements of the antenna, or on the basis of the useful signal received by these elements;
then:
in calculating the mean of the M matrices R_{xx} ^{f1} so as to obtain an autocorrelation matrix R_{xx} ^{f1} which is an estimate of E(X_{f1} (t).X_{f1} ^{+} (t)), in which E denotes the mathematical expectation, and
in calculating the mean of the M matrices R_{vv} ^{f1} so as to obtain an autocorrelation matrix R_{vv} ^{f1} which is an estimate of E(V_{f1} (t).V_{f1} ^{+} (t)), in which V_{f1} (t) is a vector having N components in which the m^{th} component represents, at the carrier frequency f_{1}, either the contribution of the interference sources and the background noise on the m^{th} element of the antenna, or the useful signal received by this element, for m varying from 1 to N.
In a second particular embodiment, employing a plurality of reception carrier frequencies f_{qr} and a plurality of transmission carrier frequencies f_{qe}, each frame of the digital signal being sent with the aid of a different carrier frequency, subject to a periodic repetition, the step of calculating the calibration tables consists furthermore in developing, for an arbitrarily chosen reception carrier frequency f_{qOr} a reception calibration table C_{qOr}, representing the variation in contribution, as a function of the reception direction, of the various reception elements at the reception carrier frequency f_{qor}.
In this second particular embodiment, the step of calculating the frequency transposition operators consists:
for each reception carrier frequency f_{qr}, in calculating a frequency transposition matrix operator T_{fqr},fqOr which transforms the reception calibration table C_{qr} associated with the frequency f_{qr} into the reception calibration table C_{qOr} associated with the frequency f_{qOr} ;
for each transmission carrier frequency f_{qe}, in calculating a frequency transposition matrix operator T_{fqOr}, fqe which transforms the calibration table C_{qOr} associated with the frequency f_{qOr} into the calibration table C_{qe} associated with the frequency f_{qe}.
Still in the second particular embodiment, the step of calculating statistical data consists for each frame of a set of K, not necessarily consecutive, frames received by the antenna, K being small enough for the angles of arrival, on the various elements of the antenna, of the multiple paths originating from the mobiles to be stable:
in calculating a matrix R_{xx} ^{fqr} =X_{fqr} (t).X_{fqr} ^{+} (t),
in which X_{fqr} (t) is a vector having N components in which the m^{th} component represents, at the carrier frequency f_{qr}, the signal received at time t from the mobiles by the m^{th} element of the antenna, for m varying from 1 to N, and
in which t describes a set of samples of the said frame, and
in estimating a matrix R_{vv} ^{fqr}, either on the basis of the contributions of the interference sources and the background noise on each of the N elements of the antenna, or on the basis of the useful signal received by these elements; then:
in calculating the mean of the K matrices R_{xx} ^{fqr} so as to obtain an autocorrelation matrix _{xx} ^{fqr} which is an estimate of E(X_{fqr} (t).X_{fqr} ^{+} (t)) and
in calculating the mean of the K matrices R_{vv} ^{fqr} so as to obtain an autocorrelation matrix R_{vv} ^{fqr} which is an estimate of E(V_{fqr} (t).V_{fqr} ^{+} (t)), in which V_{fqr} (t) is a vector having N components in which the m^{th} component represents, at the carrier frequency f_{qr}, either the contribution of the interference sources and the background noise on the m^{th} element of the antenna, or the useful signal received by this element, for m varying from 1 to N.
In the two particular embodiments above, the respective numbers of frames M and K depend, in particular, on the speed of the mobile: the greater the speed, the smaller the number of frames available. Conversely, for a mobile at socalled moderate speed, for example a cyclist or a pedestrian who is running, it will be possible to perform the calculations over a larger number of frames.
In the aforementioned first embodiment, in the presence of P mobiles containing one useful mobile with which communication is to be established, the other P1 mobiles constituting interference sources, the matrix R_{vv} ^{f1} may be established either on the basis of the contributions of the interference sources and the background noise on the various elements of the antenna, or on the basis of the useful signal received by the various elements.
When the matrix R_{vv} ^{f1} is established on the basis of the contributions of the interference sources and background noise, one possibility for estimating the matrix R_{vv} ^{f1} is that
the impulse response {g_{k},m,1, . . . ,g_{k},m,L }, in which L is an integer, of the propagation channel connecting the k^{th} mobile to the m^{th} element of the antenna, is determined for k varying from 1 to P and m varying from 1 to N;
spatial correlation matrices having N rows and ##EQU7## are formed in which the j^{th} mobile is the useful mobile,
in which G_{k},i is the column vector [g_{k},1,i, . . . ,g_{k},N,i ]^{T} and
in which (.)^{T} denotes the transpose matrix;
the mean of these spatial correlation matrices is taken over a predetermined number of frames of the signal;
for m varying from 1 to N, the variance σ_{I},m^{2} of the background noise on the m^{th} element is estimated;
the mean matrix of the spatial correlation matrices is added to a diagonal matrix in which, for m varying from 1 to N, the term located on the m^{th} row and in the m^{th} column is the variance σ_{I},m^{2}, the sum matrix obtained constituting the matrix R_{vv} ^{f1}.
For the last step, the variance σ_{I},m^{2} may be replaced by any other suitable constant.
An alternative possibility for estimating the matrix R_{vv} ^{f1}, still when it is established on the basis of the contributions of the interference sources and the background noise on the various elements of the antenna, is that
with the aid of a learning sequence of L_{ref} samples, for m varying from 1 to N, the impulse response of the propagation channel connecting the mobile to the m^{th} element of the antenna is estimated in terms of the least squares approximation, the residue b_{m} of this estimation being a column vector constituting an estimate of the contribution of the interference sources and the background noise on the m^{th} element;
the matrix B having the N vectors b_{m} ^{T} as rows is formed;
the expression (1/L_{B}).B.B^{+} is calculated, in which L_{B} denotes the number of columns of the matrix B, the matrix obtained constituting the matrix R_{vv} ^{f1}.
When the matrix R_{vv} ^{f1} is established on the basis of the useful signal received by the various elements of the antenna, on e possibility for estimating R_{vv} ^{f1} is that
the impulse response {g_{m},1, . . . ,g_{m},L }, in which L is an integer, of the propagation channel connecting the mobile to the m^{th} element of the antenna is determined for m varying from 1 to N;
spatial correlation matrices having N rows and N columns ##EQU8## are formed in which G_{i} is the column vector [g_{1},i, . . . ,g_{N},i ]^{T} ;
the mean of these spatial correlation matrices is taken over a predetermined number of frames of the signal, the matrix obtained constituting the matrix R_{vv} ^{f1}.
An alternative possibility for estimating the matrix R_{vv} ^{f1}, still when it is established on the basis of the useful signal received by the various elements of the antenna, is that
with the aid of a learning sequence of L_{ref} samples, for m varying from 1 to N, the impulse response {g_{m},1, . . . ,g_{m},L }, in which L is an integer, of the propagation channel connecting the mobile to the m^{th} element of the antenna is estimated in terms of the least squares approximation;
spatial correlation matrices having N rows and N columns ##EQU9## are formed in which G_{i} is the column vector [g_{1},i, . . . ,g_{N},i ]^{T} ;
the mean of these spatial correlation matrices is taken over a predetermined number of frames of the signal, the matrix obtained constituting the matrix R_{vv} ^{f1}.
Similarly, in the aforementioned second embodiment, in the presence of P mobiles containing a useful mobile with which communication is to be established, the other P1 mobiles constituting the interference sources, the matrix R_{vv} ^{fqr} may be established either on the basis of the contributions of the interference sources and the background noise on the various elements of the antenna, or on the basis of the useful signal received by the various elements.
When the matrix R_{vv} ^{fqr} is established on the basis of the contributions of the interference sources and the background noise, one possibility for estimating the matrix R_{vv} ^{fqr} is that
the impulse response {g_{k},m,1, . . . ,g_{k},m,L }, in which L is an integer, of the propagation channel connecting the k^{th} mobile to the m^{th} element of the antenna, is determined for k varying from 1 to P and m varying from 1 to N;
spatial correlation matrices having N rows and N columns ##EQU10## are formed in which the j^{th} mobile is the useful mobile,
in which G_{k},i is the column vector [g_{k},1,i, . . . ,g_{k},N,i ]^{T} and
in which (.)^{T} denotes the transpose matrix;
the mean of these spatial correlation matrices is taken over a predetermined number of frames of the signal;
for m varying from 1 to N, the variance σ_{I},m^{2} of the background noise on the m^{th} element is estimated;
the mean matrix of the spatial correlation matrices is added to a diagonal matrix in which, for m varying from 1 to N, the term located on the m^{th} row and in the m^{th} column is the variance σ_{I},m^{2}, the sum matrix obtained constituting the matrix R_{vv} ^{fqr}.
For the last step, the variance σ_{I},m^{2} may be replaced by any other suitable constant.
An alternative possibility for estimating the matrix R_{vv} ^{fqr}, still when it is established on the basis of the contributions of the interference sources and the background noise on the various elements of the antenna, is that
with the aid of a learning sequence of L_{ref} samples, for m varying from 1 to N, the impulse response of the propagation channel connecting the mobile to the m^{th} element of the antenna is estimated in terms of the least squares approximation, the residue b_{m} of this estimation being a column vector constituting an estimate of the contribution of the interference sources and the background noise on the m^{th} element;
the matrix B having the N vectors b_{m} ^{T} as rows is formed;
the expression (1/L_{B}).B.B^{+} is calculated, in which L_{B} denotes the number of columns of the matrix B, the matrix obtained constituting the matrix R_{vv} ^{fqr}.
When the matrix R_{vv} ^{fqr} is established on the basis of the useful signal received by the various elements of the antenna, one possibility for estimating R_{vv} ^{fqr} is that
the impulse response {g_{m},1, . . . ,g_{m},L }, in which L is an integer, of the propagation channel connecting the mobile to the m^{th} element of the antenna is determined for m varying from 1 to N;
spatial correlation matrices having N rows and N columns ##EQU11## are formed in which G_{i} is the column vector [g_{1},i, . . . ,g_{N},i ]^{T} ;
the mean of these spatial correlation matrices is taken over a predetermined number of frames of the signal, the matrix obtained constituting the matrix R_{vv} ^{fqr}.
An alternative possibility for estimating the matrix R_{vv} ^{fqr}, still when it is established on the basis of the useful signal received by the various elements of the antenna, is that
with the aid of a learning sequence of L_{ref} samples, for m varying from 1 to N, the impulse response {g_{m},1, . . . ,g_{m},L }, in which L is an integer, of the propagation channel connecting the mobile to the m^{th} element of the antenna is estimated in terms of the least squares approximation;
spatial correlation matrices having N rows and N columns ##EQU12## are formed in which G_{i} is the column vector [g_{1},i, . . . ,g_{N},i ]^{T} ;
the mean of these spatial correlation matrices is taken over a predetermined number of frames of the signal, the matrix obtained constituting the matrix R_{vv} ^{fqr}.
As indicated by the FIGURE, the next step in the method according to the present invention consists in calculating, for the useful mobile, an optimum set of spatial weightings on the basis of the statistical data and the frequency transposition operator or operators obtained previously, and on the basis of a criterion for reinforcing the useful signal and reducing the interference sources.
Next, the contributions to the digital signal to be transmitted by each element are respectively weighted by the weightings obtained on the basis of the spatial weightings of the optimum set.
Finally, the digital signal thus weighted is transmitted.
In the aforementioned first particular embodiment, the three steps which have just been described (namely calculating an optimum set of spatial weightings, weighting the signal to be transmitted and transmitting the weighted signal) may be carried out as follows:
a spatial weighting vector w_{f1} is calculated so that the matrices R_{xx} ^{f1}, R_{vv} ^{f1} and the vector w_{f1} satisfy a suitable criterion for reinforcing the useful signal and reducing the interference sources;
the inverse, denoted T_{f1},f2^{1}, of the frequency transposition operator is applied to the weighting vector w_{f1} so as to obtain the optimum set of spatial weightings in the form of a vector w_{f2} =T_{f1},f2^{1}.W_{f1} ;
for m varying from 1 to N, the signal to be transmitted by the m^{th} element of the antenna to the mobile is multiplied by the m^{th} component of the conjugate transposed optimum weighting vector W_{f2} ^{+}.
In this embodiment, when the autocorrelation matrix R_{vv} ^{f1} is established on the basis of the contributions of the interference sources and the background noise on the various elements of the antenna, the criterion consists in choosing for the weighting vector w_{f1} the vector w_{f1} which minimizes the ratio (w_{f1} ^{+}.R_{vv} ^{f1}.w_{f1})/(w_{f1} ^{+}. R_{xx} ^{f1}.w_{f1}), and when the autocorrelation matrix R_{vv} ^{f1} is established on the basis of the useful signal received by the various elements of the antenna, the criterion consists in choosing for the weighting vector w_{f1} the vector w_{f1} which maximizes the ratio (w_{f1} ^{+}.R_{vv} ^{f1}.w_{f1})/(w_{f1} ^{+}.R_{xx} ^{f1}.w_{f1}).
As a variant, in the first embodiment, the same three steps may be carried out as follows:
the frequency transposition operator T_{f1},f2 is applied to the matrix R_{xx} ^{f1} so as to obtain a matrix )R_{xx} ^{f2} =T_{f1},f2.R_{xx} ^{f1}.T_{f1},f2^{+} ;
the frequency transposition operator T_{f1},f2 is applied to the matrix R_{vv} ^{f1} so as to obtain a matrix R_{vv} ^{f2} =T_{f1},f2.R_{vv} ^{f1}.T_{f1},f2^{+} ;
the optimum set of spatial weightings is calculated in the form of a vector w_{f2} so that the matrices R_{xx} ^{f2}, R_{vv} ^{f2} and the vector w_{f2} satisfy a suitable criterion for reinforcing the useful signal and reducing the interference sources;
for m varying from 1 to N, the signal to be transmitted by the m^{th} element of the antenna to the mobile is multiplied by the m^{th} component of the conjugate transposed optimum weighting vector w_{f2} ^{+}.
In this variant, when the autocorrelation matrix R_{vv} ^{f1} is established on the basis of the contributions of the interference sources and the background noise on the various elements of the antenna, the criterion consists in choosing for the optimum weighting vector w_{f2} the vector w_{f2} which minimizes the ratio (w_{f2} ^{+}.R_{vv} ^{f2}.w_{f2})/(w_{f2} ^{+}.R_{xx} ^{f2}.w_{f2}), and when the autocorrelation matrix R_{vv} ^{f1} is established on the basis of the useful signal received by the various elements of the antenna, the criterion consists in choosing for the optimum weighting vector w_{f2} the vector w_{f2} which maximizes the ratio (w_{f2} ^{+}.R_{vv} ^{f2}.w_{f2})/(w_{f2} ^{+}.R_{xx} ^{f2}.w_{f2}).
In the aforementioned second particular embodiment, the same three steps may be carried out as follows:
for the reception carrier frequency f_{qOr}, a spatial weighting vector w_{fqOr} is calculated so that the matrices R_{xx} ^{fqOr}, R_{vv} ^{fqOr} and the vector w_{fqOr} satisfy a suitable criterion for reinforcing the useful signal and reducing the interference sources;
for each transmission carrier frequency f_{qe}, the inverse, denoted T_{fqOr},fqe^{1}, of the frequency transposition operator is applied to the weighting vector w_{fqOr} so as to obtain the optimum set of spatial weightings in the form of a vector w_{fqe} =T_{fqOr},fqe^{1}.w_{fqOr} ;
for m varying from 1 to N, the signal to be transmitted at the transmission carrier frequency f_{qe} by the m^{th} element of the antenna to the mobile is multiplied by the m^{th} component of the conjugate transposed optimum weighting vector w_{fqe} ^{+}.
In this embodiment, when the autocorrelation matrix R_{vv} ^{fqr} is established on the basis of the contributions of the interference sources and the background noise on the various elements of the antenna, the criterion consists in choosing for the weighting vector w_{fqOr} the vector w_{fqOr} which minimizes the ratio (w_{fqOr} ^{+}.R_{vv} ^{fqOr}.w_{fqOr})/(w_{fqOr} ^{+}.R_{xx} ^{fqOr}.w_{fqOr}), and when the autocorrelation matrix R_{vv} ^{fqr} is established on the basis of the useful signal received by the various elements of the antenna, the criterion consists in choosing for the optimum weighting vector w_{fqOr} the vector w_{fqOr} which maximizes the ratio (w_{fqOr} ^{+}.R_{vv} ^{fqOr}.w_{fqOr})/(w_{fqOr} ^{+}.R_{xx} ^{fqOr}.w_{fqOr}).
As a variant, in the second embodiment, the same three steps may be carried out as follows:
the K corresponding frequency transposition operators T_{fqr},fqOr are applied respectively to the K matrices R_{xx} ^{fqr} so as to obtain K matrices R_{xx} ^{fqOr} =T_{fqr},fqOr.R_{xx} ^{fqr}.T_{fqr},fqOr^{+} ;
the K frequency transposition operators T_{fqr},fqOr are applied respectively to the K matrices R_{vv} ^{fqr} so as to obtain K matrices R_{vv} ^{fqOr} =T_{fqr},fqOr.R_{vv} ^{fqr}.T_{fqr},fqOr^{+} ;
the mean R_{xx} ^{fqOr} of the K matrices R_{xx} ^{fqOr} and the mean R_{vv} ^{fqOr} of the K matrices R_{vv} ^{fqOr} are calculated;
for each transmission carrier frequency f_{qe}, the frequency transposition operator T_{fqOr},fqe is applied to the mean matrix R_{vv} ^{fqOr} so as to obtain a matrix R_{xx} ^{fqe} =T_{fqOr},fqe.R_{xx} ^{fqOr}.T_{fqOr},fqe^{+} ;
for each transmission carrier frequency f_{qe}, the frequency transposition operator T_{fqOr},fqe is applied to the mean matrix R_{vv} ^{fqOr} so as to obtain a matrix R_{vv} ^{fqe} =T_{fqOr},fqe.R_{vv} ^{fqOr}.T_{fqOr},fqe^{+} ;
for each transmission carrier frequency f_{qe}, the optimum set of spatial weightings is calculated in the form of a vector w_{fqe} so that the matrices R_{xx} ^{fqe}, R_{vv} ^{fqe} and the vector w_{fqe} satisfy a suitable criterion for reinforcing the useful signal and reducing the interference sources;
for m varying from 1 to N, the signal to be transmitted at the transmission carrier frequency f_{qe} by the m^{th} element of the antenna to the mobile is multiplied by the m^{th} component of the conjugate transposed optimum weighting vector w_{fqe} ^{+}.
In this variant, when the autocorrelation matrix R_{vv} ^{fqr} is established on the basis of the contributions of the interference sources and the background noise on the various elements of the antenna, the criterion consists in choosing for the optimum weighting vector w_{fqe} the vector w_{fqe} which minimizes the ratio (w_{fqe} ^{+}.R_{vv} ^{fqe}.w_{fqe})/(w_{fqe} ^{+}.R_{xx}.sup.fqe.w_{fqe}), and when the autocorrelation matrix R_{vv} ^{fqr} is established on the basis of the useful signal received by the various elements of the antenna, the criterion consists in choosing for the optimum weighting vector w_{fqe} the vector w_{fqe} which maximizes the ratio (w_{fqe} ^{+}.R_{vv} ^{fqe}.w_{fqe})/(w_{fqe} ^{+}.R_{xx}.sup.fqe.w_{fqe}).
In yet another embodiment, the frequency transposition operator (step c of the method) is implicit. The approach is then as follows. A set of weightings w_{fqr} which is optimum at the reception frequency is determined. The set of weightings at the transmission frequency is then calculated by adjusting, in terms of least squares, the antenna diagram W*_{fqe} S.sub.λe (α) at f_{qe} to the antenna diagram W*_{fqr} S.sub.λ (α) at the reception frequency f_{qr}.
The weightings thus obtained are equal to T*_{fqefqr} W_{fqr} in which T_{fqefqr} is the frequency transposition matrix from f_{qe} to f_{qr}.
From the point of view of implementation means, functions and results, a variant of this type is fully equivalent to the one defined in detail above.
Claims (27)
Priority Applications (2)
Application Number  Priority Date  Filing Date  Title 

FR9613597A FR2755565B1 (en)  19961107  19961107  A method of transmitting by a base station equipped with an antenna has a plurality of sensors to a mobile 
FR9613597  19961107 
Publications (1)
Publication Number  Publication Date 

US6021334A true US6021334A (en)  20000201 
Family
ID=9497424
Family Applications (1)
Application Number  Title  Priority Date  Filing Date 

US08965501 Expired  Lifetime US6021334A (en)  19961107  19971106  Method for transmission by a base station equipped with a multielement antenna to a mobile 
Country Status (6)
Country  Link 

US (1)  US6021334A (en) 
EP (1)  EP0841759B1 (en) 
JP (1)  JP4087483B2 (en) 
DE (2)  DE69732335T2 (en) 
ES (1)  ES2235221T3 (en) 
FR (1)  FR2755565B1 (en) 
Cited By (8)
Publication number  Priority date  Publication date  Assignee  Title 

US6346910B1 (en)  19990407  20020212  Tei Ito  Automatic array calibration scheme for wireless pointtomultipoint communication networks 
US20020057660A1 (en) *  20001012  20020516  Hyung Gun Park  Method and apparatus for searcher beamforming in CDMA base station system using array antenna 
EP1317080A1 (en) *  20011203  20030604  Alcatel  An apparatus and method of transmitting signals between an adaptive antenna of a base station and a mobile user equipment in a telecommunication network 
US20030228017A1 (en) *  20020422  20031211  Beadle Edward Ray  Method and system for waveform independent covert communications 
US20030231606A1 (en) *  20010720  20031218  Hebing Wu  Method and apparatus for downlink feedback multiple antenna transmission in wireless communication system 
US20040189525A1 (en) *  20030328  20040930  Beadle Edward R.  System and method for cumulantbased geolocation of cooperative and noncooperative RF transmitters 
US20070104152A1 (en) *  20051104  20070510  Alcatel  Method for preforming user allocation in SDMA systems, and corresponding base station 
US20080049647A1 (en) *  19991209  20080228  Broadcom Corporation  Voiceactivity detection based on farend and nearend statistics 
Families Citing this family (2)
Publication number  Priority date  Publication date  Assignee  Title 

FR2766627B1 (en) *  19970728  19991001  France Telecom  Antenna array for base station radio communication with mobile 
WO2001017065A1 (en) *  19990827  20010308  Telefonaktiebolaget Lm Ericsson (Publ)  Method for electronical beam control in a telecommunications system and base station using said method 
Citations (10)
Publication number  Priority date  Publication date  Assignee  Title 

US5274844A (en) *  19920511  19931228  Motorola, Inc.  Beam pattern equalization method for an adaptive array 
EP0639035A1 (en) *  19930812  19950215  Northern Telecom Limited  Base station antenna arrangement 
US5515378A (en) *  19911212  19960507  Arraycomm, Inc.  Spatial division multiple access wireless communication systems 
US5546090A (en) *  19911212  19960813  Arraycomm, Inc.  Method and apparatus for calibrating antenna arrays 
US5557640A (en) *  19930716  19960917  Plessey Semiconductor Limited  Phase and amplitude equalization arrangement 
US5710995A (en) *  19970116  19980120  Ford Motor Company  Adaptive antenna receiver 
US5875216A (en) *  19960227  19990223  Lucent Technologies Inc.  Weight generation in stationary interference and noise environments 
US5887037A (en) *  19960227  19990323  Lucent Technologies Inc.  Introducing processing delay as a multiple of the time slot duration 
US5893033A (en) *  19950524  19990406  Nokia Telecommunications: Oy  Methods for making a faster handover, and a cellular radio systems 
US5930243A (en) *  19961011  19990727  Arraycomm, Inc.  Method and apparatus for estimating parameters of a communication system using antenna arrays and spatial processing 
Patent Citations (10)
Publication number  Priority date  Publication date  Assignee  Title 

US5515378A (en) *  19911212  19960507  Arraycomm, Inc.  Spatial division multiple access wireless communication systems 
US5546090A (en) *  19911212  19960813  Arraycomm, Inc.  Method and apparatus for calibrating antenna arrays 
US5274844A (en) *  19920511  19931228  Motorola, Inc.  Beam pattern equalization method for an adaptive array 
US5557640A (en) *  19930716  19960917  Plessey Semiconductor Limited  Phase and amplitude equalization arrangement 
EP0639035A1 (en) *  19930812  19950215  Northern Telecom Limited  Base station antenna arrangement 
US5893033A (en) *  19950524  19990406  Nokia Telecommunications: Oy  Methods for making a faster handover, and a cellular radio systems 
US5875216A (en) *  19960227  19990223  Lucent Technologies Inc.  Weight generation in stationary interference and noise environments 
US5887037A (en) *  19960227  19990323  Lucent Technologies Inc.  Introducing processing delay as a multiple of the time slot duration 
US5930243A (en) *  19961011  19990727  Arraycomm, Inc.  Method and apparatus for estimating parameters of a communication system using antenna arrays and spatial processing 
US5710995A (en) *  19970116  19980120  Ford Motor Company  Adaptive antenna receiver 
NonPatent Citations (3)
Title 

Adaptive Transmitting Antenna Methods for Multipath Environments, Derek Gerlach and Arogyaswami Paulraj Globecom 1994, pp. 425 429. * 
Adaptive Transmitting Antenna Methods for Multipath Environments, Derek Gerlach and Arogyaswami Paulraj Globecom 1994, pp. 425429. 
French Search Report dated Jul. 18, 1997, French Appl No. FR 9613597 filed Nov. 7, 1996. * 
Cited By (16)
Publication number  Priority date  Publication date  Assignee  Title 

US6346910B1 (en)  19990407  20020212  Tei Ito  Automatic array calibration scheme for wireless pointtomultipoint communication networks 
US20110058496A1 (en) *  19991209  20110310  Leblanc Wilfrid  Voiceactivity detection based on farend and nearend statistics 
US7835311B2 (en) *  19991209  20101116  Broadcom Corporation  Voiceactivity detection based on farend and nearend statistics 
US20080049647A1 (en) *  19991209  20080228  Broadcom Corporation  Voiceactivity detection based on farend and nearend statistics 
US8565127B2 (en)  19991209  20131022  Broadcom Corporation  Voiceactivity detection based on farend and nearend statistics 
US7027421B2 (en)  20001012  20060411  Electronics And Telecommunications Research Institute  Method and apparatus for searcher beamforming in CDMA base station system using array antenna 
US20020057660A1 (en) *  20001012  20020516  Hyung Gun Park  Method and apparatus for searcher beamforming in CDMA base station system using array antenna 
US20030231606A1 (en) *  20010720  20031218  Hebing Wu  Method and apparatus for downlink feedback multiple antenna transmission in wireless communication system 
US7236478B2 (en) *  20010720  20070626  Huawei Technologies Co., Ltd.  Method and apparatus for downlink feedback multiple antenna transmission in wireless communication system 
US20030104837A1 (en) *  20011203  20030605  Alcatel  Method of transmitting signals between an adaptive antenna of a base station and a mobile user equipment in a telecommunication network 
US7239848B2 (en)  20011203  20070703  Alcatel  Method of transmitting signals between an adaptive antenna of a base station and a mobile user equipment in a telecommunication network 
EP1317080A1 (en) *  20011203  20030604  Alcatel  An apparatus and method of transmitting signals between an adaptive antenna of a base station and a mobile user equipment in a telecommunication network 
US20030228017A1 (en) *  20020422  20031211  Beadle Edward Ray  Method and system for waveform independent covert communications 
US7187326B2 (en) *  20030328  20070306  Harris Corporation  System and method for cumulantbased geolocation of cooperative and noncooperative RF transmitters 
US20040189525A1 (en) *  20030328  20040930  Beadle Edward R.  System and method for cumulantbased geolocation of cooperative and noncooperative RF transmitters 
US20070104152A1 (en) *  20051104  20070510  Alcatel  Method for preforming user allocation in SDMA systems, and corresponding base station 
Also Published As
Publication number  Publication date  Type 

FR2755565B1 (en)  19990108  grant 
EP0841759A1 (en)  19980513  application 
JPH10247875A (en)  19980914  application 
EP0841759B1 (en)  20050126  grant 
ES2235221T3 (en)  20050701  grant 
JP4087483B2 (en)  20080521  grant 
DE69732335D1 (en)  20050303  grant 
FR2755565A1 (en)  19980507  application 
DE69732335T2 (en)  20051027  grant 
Similar Documents
Publication  Publication Date  Title 

US5875216A (en)  Weight generation in stationary interference and noise environments  
Razavilar et al.  Software radio architecture with smart antennas: a tutorial on algorithms and complexity  
US6654590B2 (en)  Determining a calibration function using at least one remote terminal  
US6512737B1 (en)  Stacked carrier discrete multiple tone communication system  
US7139592B2 (en)  Null deepening for an adaptive antenna based communication station  
Paulraj et al.  Introduction to spacetime wireless communications  
Wu et al.  Blind adaptive beamforming for cyclostationary signals  
US6101399A (en)  Adaptive beam forming for transmitter operation in a wireless communication system  
US20040235421A1 (en)  Radio communication apparatus using adaptive antenna  
US20080181174A1 (en)  Method and apparatus for a transceiver in a mobile/fixed relay with multiple antennas  
US6331837B1 (en)  Spatial interferometry multiplexing in wireless communications  
Blanz et al.  Smart antennas for combined DOA and joint channel estimation in timeslotted CDMA mobile radio systems with joint detection  
US7020490B2 (en)  Radio communication system  
US6754511B1 (en)  Linear signal separation using polarization diversity  
US7385617B2 (en)  Methods for multiuser broadband wireless channel estimation  
US6141393A (en)  Method and device for channel estimation, equalization, and interference suppression  
US6219561B1 (en)  Wireless communication network using timevarying vector channel equalization for adaptive spatial equalization  
US20040012387A1 (en)  Cancellation systems for multicarrier transceiver arrays  
US6901122B2 (en)  Method and apparatus for restoring a soft decision component of a signal  
US20100087151A1 (en)  Radio Channel Estimator  
US7548592B2 (en)  Multiple input, multiple output communications systems  
US20030012308A1 (en)  Adaptive channel estimation for wireless systems  
US20100158151A1 (en)  System and method for combination multiple input, multiple output (mimo) and beamforming  
US6590532B1 (en)  Radio device  
EP1276251A1 (en)  Method for calculating a weighting vector for an antenna array 
Legal Events
Date  Code  Title  Description 

AS  Assignment 
Owner name: FRANCE TELECOM, FRANCE Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:ASTE, THIERRY;FETY, LUE;FORSTER, PHILIPPE;AND OTHERS;REEL/FRAME:009043/0509;SIGNING DATES FROM 19971117 TO 19971127 Owner name: FRANCE TELECOM, FRANCE Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:ASTE, THIERRY;FETY, LUE;FORSTER, PHILIPPE;AND OTHERS;SIGNING DATES FROM 19971117 TO 19971127;REEL/FRAME:009043/0509 

FPAY  Fee payment 
Year of fee payment: 4 

FPAY  Fee payment 
Year of fee payment: 8 

AS  Assignment 
Owner name: FRANCE TELECOM, FRANCE Free format text: CHANGE OF LEGAL STATUS OF FRENCH COMPANY TO SOCIETE ANONYME;ASSIGNOR:FRANCE TELECOM;REEL/FRAME:022052/0983 Effective date: 19970109 

AS  Assignment 
Owner name: GULA CONSULTING LIMITED LIABILITY COMPANY, DELAWAR Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:FRANCE TELECOM SA;REEL/FRAME:022354/0124 Effective date: 20081202 

FPAY  Fee payment 
Year of fee payment: 12 